A curated, graph-based living review of computational methods and datasets for spatial omics analysis. Built out of frustration with static reviews that become outdated before they are published.
The spatial omics methods landscape moves faster than any static review can keep up with. SOLR is an attempt to fix that — a living, browsable graph of computational methods, benchmarking relationships, and datasets, maintained by someone who actually used these tools during a PhD.
The core idea: methods are only meaningful in context. Which datasets were used to benchmark them? Against which other methods? A tool validated only on mouse brain cortex is a different thing from one validated on human tumour tissue. SOLR makes that context visible.
- Methods graph — computational methods grouped by pipeline stage (preprocessing, cell segmentation, phenotyping, niche analysis, spatially variable genes, cell-cell communication), with edges showing which methods were compared against each other and on what data
- Dataset registry — curated spatial omics datasets with tissue type, disease, organism, modality, technology, and number of markers/genes — with links to download, nothing hosted here
- Methods book — a narrative overview of the methods landscape, written during a PhD in spatial omics
Phase 1 — manually curated static version. A core set of well-curated papers with real comparison edges, the rest as stubs. Hosted as static files on GitHub Pages, no backend.
Phase 2 will introduce an automated review pipeline — a set of agents that scrape bioRxiv and PubMed, categorise new papers, extract which methods they compare against and which datasets they use, and populate the database automatically. The goal is a graph that stays current without manual effort, while preserving manual curation quality markers so users know what to trust.
data/
data_curated/ ← source of truth (Excel)
data_curated_backup/ ← versioned backups
processed/ ← cleaned CSVs and JSON outputs
src/
preprocessing/
01_parse_excel.py ← clean and export from Excel
02_fetch_metadata.py ← enrich with Crossref/PubMed metadata
03_export_json.py ← export to JSON for the frontend
docs/ ← static site (GitHub Pages)
index.html
data/
book/
version1/ ← archived first prototype
If you want to suggest a missing method, flag an error, or add a dataset — open an issue or a pull request. The source of truth is the curated spreadsheet; contributions to that are the most valuable.
Started during a PhD in spatial omics. The man in the boat knows the feeling.
